Radiation protection measurements around a 12 MeV mobile dedicatedIORT accelerator
Antonella Soriania�
Laboratory of Medical Physics, Istituto Regina Elena, via Elio Chianesi 53, 00144 Rome, Italy
Giuseppe Felici and Mario FantiniSordina SpA Technical Division, via Calanna 25, 00126 Rome, Italy
Massimiliano PaolucciDepartment of Medical Physics, ASL n.3, Umbria V. Arcamone 1, 06034 Foligno (Perugia), Italy
Oscar BorlaI.N.F.N. sez. Torino, Via Pietro Giuria 1, 10125 Torino, Italy
Giovanna Evangelisti, Marcello Benassi, and Lidia StrigariLaboratory of Medical Physics, Istituto Regina Elena, via Elio Chianesi 53, 00144 Roma, Italy
�Received 4 May 2009; revised 24 December 2009; accepted for publication 30 December 2009;published 5 February 2010�
Purpose: The aim of this study is to investigate radioprotection issues that must be addressed whendedicated accelerators for intraoperative radiotherapy �IORT� are used in operating rooms. Re-cently, a new version of a mobile IORT accelerator �LIAC™ Sordina SpA, Italy� with 12 MeVelectron beam has been implemented. This energy is necessary in some specific pathology treat-ments to allow a better coverage of thick lesions. At an electron energy of 10 MeV, leakage andscattered x-ray radiation �stray radiation� coming from the accelerator device and patient must beconsidered. If the energy is greater than 10 MeV, the x-ray component will increase; however, themost meaningful change should be the addition of neutron background. Therefore, radiation expo-sure of personnel during the IORT procedure needs to be carefully evaluated.Methods: In this study, stray x-ray radiation was measured and characterized in a series of spheri-cal projections by means of an ion chamber survey meter. To simulate the patient during allmeasurements, a polymethylmethacrylate �PMMA� slab phantom with volume 30�30�15 cm3
Key words: IORT, linear accelerators, electron beams, neutron contamination, radiation protection
I. INTRODUCTION
The term intraoperative radiotherapy �IORT� refers to theapplication of radiation in a single session either after surgi-cal removal of neoplastic tissue or not, as the case may be. Inthe past, IORT was carried out using electron beams pro-duced by conventional linear accelerators, transferring thepatient from the operating theater to a shielded radiotherapy
department, and then back again.
995 Med. Phys. 37 „3…, March 2010 0094-2405/2010/37„3
In the past ten years, there has been an increasing interestin the IORT technique also because of the development ofmobile accelerators1–3 producing only electron beams. Thistype of machine can be introduced directly into an operatingtheater with no need for any special fixed shielding barriers.Mobile machines solve logistical problems, such as the needfor transporting the anesthetized patient, thereby reducingthe overall time of the procedure.
The photoneutron contamination arising from conven-
996 Soriani et al.: Radiation protection around 12 MeV dedicated IORT accelerator 996
tional and IORT dedicated accelerators has been studied bymany authors.4–7 Photoneutrons are mostly produced in theaccelerator head and a lower amount comes from the patient,phantom, or shield. According to the principles of radiationprotection, the clinical use of electron linear accelerators re-quires an accurate study of the radiation source and shieldingto ensure that radiation exposure limits defined by the Euro-pean Community for radiation protection of workers andpopulation are met.8
In the framework of a national project on quality assur-ance in Radiotherapy,9 the Italian National Institute of Healthhas established a multidisciplinary working group, with ex-perience in clinical practice, to develop quality assuranceguidelines for the IORT technique. In this context, new re-quirements have been discussed regarding the need for tech-nological innovations and for the development of more ef-fective accelerator models in order to allow new clinicalapplications �such as breast cancer or sterilization of thicklesions, as prostate�.
Thus, some IORT mobile accelerators are increasing themaximum electron beam energy over 10 MeV to treat largertumors. While energies up to 10 MeV pose no specialneutron-shielding problems,3 neutron measurements must tobe made for higher energy beams. In fact, in clinical practice,the therapeutic range of 12 MeV could be helpful in thetreatment of specific pathologies.10
To increase the beam energy of the mobile IORT accel-erator, named LIAC™, Sordina SpA �Italy�11 developed anew model to restyle the linear accelerator structure and in-troduced severe dosimetric controls to check the perfor-mance of this new machine. During the new accelerator de-velopment phase and assembly, a close cooperation betweenthe manufacturer and the customer has led to improvedchoices of materials and design. In this study, the aim is toreport on the neutron and x-ray radiation produced by thenew LIAC™.
II. MATERIALS AND METHODS
II.A. IORT accelerator characteristics
The light intraoperative accelerator �LIAC™� is a mobileaccelerator specifically designed to perform IORT in operat-ing theaters. The LIAC™ consists of a mobile linear accel-erator and an operator control rack connected by a cable. Themobile unit weighs about 400 kg, while the control rackweighs less than 100 kg. During the irradiation, the LIAC™is not connected to the electrical local system but runs onbatteries �uninterruptible power supply�. An innovative ro-botic system allows the LIAC™ to be extremely mobile andstrongly simplifies hard-docking procedures. The LIAC™head has three degrees of freedom: It can be moved up anddown for a maximum excursion of 100 cm, it has a roll angleof �60°, and a pitch angle between +30° and �15°.
The 10 MeV model has been on the market since 2001and, to meet customer demands, a new LIAC™ model ableto accelerate electrons up to 12 MeV was recently developed.
The accelerating system is a newly designed linac operat-
ing in the � /2 mode at 2998 MHz �S band�. The 12 MeV
Medical Physics, Vol. 37, No. 3, March 2010
linac is 92.5 cm long �19 accelerating cavities� and its totalweight, including electron gun and ionic vacuum pumps, isless than 30 kg. Radiofrequency power is supplied by anE2V magnetron MG6090. The electron energy is set by vary-ing the radiofrequency power from 1.2 up to 3 MW. The newmachine setup provides four clinical energy points: 6, 8, 10,and 12 MeV. It was manufactured in accordance with Italianregulations.11
The particular architecture of the LIAC™ acceleratorhead guarantees a minimal head leakage radiation, muchlower than target scatter radiation. The accelerating wave-guide has no external solenoid for electron beam radial fo-cusing, but electrostatic focusing is used instead.12 This ra-dial focusing system decreases the tail of electron beamdistribution hitting the copper waveguide, reducing brems-strahlung radiation, and focuses the electrons along the beamline.
Furthermore, there is no bending magnet, and the metallicelements which the electron beam crosses along its path are atitanium window, 55 �m thick, an aluminum scattering foil,820 �m thick, and four ionization chamber electrodes, to-tally 20 �m thick. The total head leakage, integrated overthe solid angle, is less than the scatter radiation at 90° by afactor of 10.
The choice of using an aluminum scattering foil�820 �m� for the 12 MeV instead of a brass one �75 �m�,as in the previous model �10 MeV�, was made after an ex-perimental study where the optimization parameters had thefollowing characteristics: Limited applicator length �formanageability reasons�, beam flatness �within �5% evalu-ated at 80% of the dose profile�, a controlled environmentalx-ray radiation, and a low neutron contamination.
The polymethylmethacrylate �PMMA� applicators are 60cm long and 0.5 cm thick and fully gas sterilizable; variousdiameters �from 30 to 100 mm typically� and bevel anglesare available. The distance from the scattering foil to the endof the applicator or source surface distance �SSD� is 713mm.
To highlight these distinctive features related to radio-therapy applications, LIAC™ 12 MeV depth dose curves ob-tained with a 100 mm applicator and in the absence of anapplicator �open field� were measured. These measurementswere compared to those obtained from a conventional accel-erator Clinac 2100CD �Varian Medical Systems, Palo Alto,CA�, with the same nominal energy �12 MeV� and a10�10 cm2 square field.
Typical dose per pulse values are reported in Table I. It isworth noting that by reducing the applicator diameter �from100 to 40 mm�, the dose per pulse increases correspondingly.This type of collimation together with the absence of a bend-ing magnet and the planning choice of light material makethis equipment able to work in operating rooms, keeping thestray radiation at low level.
Pulse repetition frequency �PRF� can be varied from 1 to60 Hz. The PRF is set by the manufacturer according to thevarious e-beam energies to keep the dose rate around 10
Gy/min with an applicator diameter of 100 mm. However, up
997 Soriani et al.: Radiation protection around 12 MeV dedicated IORT accelerator 997
to 30 Gy/min higher or lower dose rates are readily obtained.A newly designed system was used to guarantee the LIACoutput reliability. The current injected by the electron guncan be adjusted ��5% maximum� by an automatic dose con-trol board to keep the reading of the two monitor chambersconstant, so that the ratio cGy/MU is kept reliably constant.
II.B. Dosimeters and measurements layout
Beam dosimetry and radiation protection measurementswere carried out in an open space free of obstructing objects.
Percent depth dose �PDD� measurements were carried outby means of a Scanditronix Wellhöfer water phantomequipped with solid state detectors, and also using a parallelplate chamber PPC05 �Scanditronix Wellhöfer GmbH,Schwarzenbruck, Germany�. Measurements obtained usingthese two different dosimetric systems were in very goodagreement.
Absolute dose measurements were performed using twoparallel plate chambers: PPC05 Scanditronix Wellhöfer 13
and Roos PTW �Ref. 14� in a water phantom. Absolute doseswere calculated according to the international protocol15 andksat was evaluated according to a new published method.16,17
Due to the accelerating cavity design, as previously de-scribed, scatter radiation is greater than head leakage, and themajor source of photon radiation is the patient, acting as atarget for the electron beam. To simulate the patient, aPMMA slab phantom with volume 30�30�15 cm3 anddensity 1.19 g /cm3 was used and the dosimeter was placedat the depth of maximum dose on the beam central axis,named reference point �Rmax�.
Stray radiation measurements were performed using anInovision Victoreen ion chamber survey meter model 451B�Inovision, Cleveland, OH�,18 having a volume of 349 cm3
and calibrated in dose equivalent rate units of Sievert/hourfor gamma and x rays in the energy range of 20 keV–2 MeV.The accuracy of the instrument reading is �10%. To reducethe effect of backscatter, the dosimeter was enclosed in a 5cm thick lead box with only the front of the chamber facingtoward the source. The beta radiation component coming outfrom the side of the PMMA applicator was shielded bymeans of a 2 cm PMMA in front of the open side. Attenua-tion of photons due to the PMMA thickness was consideredand the measured values was correcting for the attenuationfactor of stray radiation due to the PMMA. Scattered photonradiation at different angles was taken into account by mov-
TABLE I. The typical of dose per pulse values of 12 MeV LIAC™.
ing the survey meter around the accelerator by increasing the
Medical Physics, Vol. 37, No. 3, March 2010
angle in 15° steps always facing the reference point in thePMMA phantom �see Fig. 1�. We considered 0° as a startingangle when the chamber was on the beam central axis underthe PMMA phantom.
All measurements taken for electron beam energies avail-able on the machine 6, 8, 10, and 12 MeV were made usingthe reference applicator �100 mm diameter�. Each measure-ment was executed by irradiating the PMMA for 1 min witha dose rate of 10 Gy/min. The XZ plane is visualized in afrontal view of the LIAC. These measurements help charac-terize the radiation risk to floors above and below the unit.Measurements were also performed in the horizontal XYplane. The XY plane measurements are required to determinethe risk of personnel working near the operating room on thesame floor.
For the maximum energy �12 MeV� we have also mea-sured the half value layers �HVLs� and tenth value layers�TVLs� of the x-ray component at different angles 0°, 90°,and 180°. Attenuation measurements were performed from 0to 5 cm using lead sheets of 5 mm thickness.
Radioprotection measurements were also repeated using amobile beam stopper to shield the x-ray component along thebeam axis. It consisted of 20 mm of polyethylene, 13 mm ofsteel, 40 mm of paraffin wax, and 100 mm of lead, with anarea of 45�45 cm2, mounted on a cart with an adjustableheight ranging from 55 to 80 cm. This is the standard mobilebeam stopper provided by manufacturer with LIAC™, bothin the 10 and in the 12 MeV model; in case of an eventualvery high workload �more than 300 Gy/week with 12 MeVelectron beam�, additional lead sheets can be added to thebeam absorber in order to increase its attenuation. Anothermobile lateral shield for x-ray stray radiation, made of 53mm of steel and 10 mm of polycarbonate, with an area of60�60 cm2, was used to reduce exposure at 90° with re-
FIG. 1. Experimental setup reporting position of ionization chamber duringradioprotection measurements.
spect to the beam axis. During the measurements, the lateral
998 Soriani et al.: Radiation protection around 12 MeV dedicated IORT accelerator 998
panels were positioned about 40 cm from the beam axis toreproduce the real position near the surgical bed in the oper-ating room.
Interaction of the electron beam with matter can give riseto secondary radiation that consist of photons produced bybremsstrahlung induced in various targets along the beamaxis19 �such as the titanium window of electron waveguide,the aluminum scattering foil, the transmission ionizationchambers, and the patient�.
The yield ratio between bremsstrahlung induced in targetand the electron disintegration process depends on the targetthickness and on the ratio of the radiative to the total stop-ping power for the target, as well described in literature.20,3
Photonuclear reactions, such as �� ,n�, can be generated bybremsstrahlung radiation produced in the target. The photo-neutron energy threshold depends on the isotope of thetarget.21,22
The photoneutron energy spectrum is characterized by anevaporation peak in the low energy range and a relativelyweak �10% of the integrated intensity� direct-reaction com-ponent in the several MeV energy range, resulting in a meanneutron energy of 700 keV–1 MeV. At energy levels below12 MeV, we can estimate that the neutron angular distribu-tion is isotropic because neutrons produced with an angulardistribution proportional to sin2 � �� is the angle betweenphotons and neutron directions� are only a small percentageof the whole spectrum.23
To overcome the difficulties raised by an intense photonfield when measuring neutrons, as is the case in the electronlinear accelerators used for radiotherapy,23 a set of passiveneutron detectors not sensitive to x rays based on super-heated bubble detectors �BTI Bubble Technology Industries,Ontario, Canada�24 were used. The dosimeters were cali-brated at the factory in terms of equivalent dose using anAm–Be source �strength=1.13�107 n /s, fluence weighedaverage energy 4.15 MeV� according to NCRP38.25
During our measurements we used two models of thesedetectors: BD100R and BDT, which together cover a widespectrum of neutron energy.
Bubble detector BD100R is energy independent above thethreshold, dose rate independent, gives tissue-equivalentdose measurements and has an isotropic angular responsewith a neutron energy threshold of approximately 100keV.26,27
If the room temperature differs from 20 °C, a conversiongraph describing the temperature response is supplied withthe BD100R. The bubble detector is a neutron dosimeter thatgives an integral measure over time and energy. It is notsensitive to gamma radiation, it has poor spatial resolution,and uses a slow reading dosimetry method, that is, countingeach bubble one by one with the naked eye. With an isotropicangular response, neutron doses can be accurately measuredregardless of the direction of neutrons in relation to the de-tector. The BDT model is similar to BD100R one in perfor-mance but is preferentially sensitive to thermal neutrons,with an exclusion ratio of thermal-to-fast neutron response
exceeding 10:1.
Medical Physics, Vol. 37, No. 3, March 2010
The individual calibration factor of the detector used inthe experiments ranged from 0.047 bub /�Sv �low sensitiv-ity� to 4 bub /�Sv �high sensitivity� and was stated on thedetector label. The sensitivity accuracy of the bubble detec-tor reported by the manufacturer was �20% over the tem-perature range from 20 to 37 °C.
The neutron equivalent dose was also measured by a neu-tron probe �LB6411 model� manufactured by Berthold Tech-nologies �GmbH & Co. KG, Bad Wildbad, Germany�. Thisprobe consists of a specially developed 3He2 proportionalcounter tube �40 cm2�, within a 25 cm radius polyethylenemoderating sphere. Its geometrical arrangement wasachieved with the help of Monte Carlo calculations and com-prehensive calibrations with monoenergetic neutrons, result-ing in 30% energy dependence in the energy range 50keV–10 MeV. The instrument has dose equivalent conver-sion factors defined according to ICRP 60 �Ref. 28� and hasa significantly improved response and a better detection limitas compared to conventional “rem counter.” Sensitivity ofthe overall system is 3.15 counts/nSv relative to 3 MeV. Theresponse for photon radiation is�0.69�0.05�10−3 counts /nSv, for a large energy range. Theinstrument was calibrated to neutron ambient dose equivalentH��10� according to ICRP 74.29
Neutron leakage measurements were performed by posi-tioning the accelerator with the reference applicator perpen-
FIG. 2. Experimental setup reporting the measurements points of the neutronequivalent dose rate produced by LIAC™. When measurements were per-formed without the beam stopper, a spongelike support was used.
dicular to the floor. In order not to affect the neutron and
999 Soriani et al.: Radiation protection around 12 MeV dedicated IORT accelerator 999
photons measurements, the phantom used to simulate the pa-tient was placed on a table made of light material �sponge-like�.
The sketch of the experimental setup is shown in Fig. 2.The position for the absolute dose ionization chamber in thePMMA phantom was at the depth of maximum dose on thebeam central axis, Rmax. The dose delivered for each mea-surement was about 50 Gy. We performed three irradiationsfor each point. Bubbles were counted by eye by three to fiveoperators and the average was reported. The measurementstaken with the bubble detectors were carried out: �A� On theexit window and scattering foil plane; �B� on ionizationchamber plane; �C� at a point on the external surface of thePMMA applicator; on the phantom surface: �D� Outside and�E� inside the applicator; ��F� and �G�� on the border of thephantom 15 cm far from the beam axis and �L� 1 m from theapplicator axis; �H� just under the phantom; and �I� on thefloor on the central beam axis. Measurements at points B, E,and H were performed also without the PMMA applicator.
The measurements indicated in Fig. 2 as E, H, and I werealso repeated with the mobile beam stopper in its clinical useposition. The measurements with LB6411 neutron probewere carried out in the phantom plane at 1 m orthogonaldistance from the applicator axis �L� and on the floor, on thecentral beam axis, about 140 cm distant from the scatteringfoil �I� without the beam absorber shield.
III. RESULTS
Figure 3 shows the PDD carried out at 12 MeV for theLIAC™ with and without the 100 mm diameter applicator
FIG. 3. Percent depth dose curves obtained for the 12 MeV LIAC™ withou12 MeV electron beams �-�-�.
and for a 12 MeV electron beam from a conventional linac
Medical Physics, Vol. 37, No. 3, March 2010
Clinac 2100CD �Varian Medical Systems, Palo Alto, CA�.The passive beam shaping technique used by the LIAC™delivers a quite uniform and flat radiation field �specifically�3% for applicators with diameters greater than 50 mm and�5% for smaller applicators� allowing very low x-ray con-tamination. Furthermore, the interaction between the electronbeam and the PMMA applicator generates low energy elec-trons which deposit dose in the region very close to the sur-face. This fact explains the higher value of the skin dose incomparison with the conventional linac. The main observa-tion is that with the reference applicator, the surface dose is94% of the maximum dose and Rmax is decreased. Moreover,Fig. 3 points out that the x-ray contamination of the 12 MeVLIAC™, with or without applicator, is only 0.6% of themaximum dose compared 2.5% of the maximum dose of aconventional linear accelerator. We can also note that thedepth of d90 �90% of dose� is reduced by 10 mm from theconventional linac to the IORT case with the applicator and 5mm from the conventional linac to the no applicator IORTcase.
The beam quality for the leakage photon and scattered
TABLE II. HVL and TVL at various angles for 12 MeV LIAC™.
licator �-�-�, with a 100 mm diameter applicator �-�-�, and for the Clinac
t app
1000 Soriani et al.: Radiation protection around 12 MeV dedicated IORT accelerator 1000
component has been measured at 0° �along the beam axis�and at 90° with respect to the beam central axis �see point L,Fig. 2� and is reported in Table II in terms of TVL and HVL.Data refer to measured for lead values, and calculated forordinary concrete values. The backscatter component mea-
FIG. 4. Stray radiation dose equivalent per unit of electron beam dose de-livered at Rmax ��Sv /Gy�, 12 MeV beam energy, 100 mm diameter appli-cator, SSD 60 cm, 1 m away from electron beam impact on patient in the XZplane �see Fig. 1� �a� with and without the beam stopper; and �b� with andwithout the lateral shield and beam stopper.
sured at 180° was negligible.
Medical Physics, Vol. 37, No. 3, March 2010
The x-ray stray radiation curves, for 1 Gy at Rmax deliv-ered using the 12 MeV LIAC™ with the reference applica-tor, were measured at 1 m and at different angles �Fig. 4�a��with and without the mobile beam stopper. The x-ray strayradiation curves were also measured as already described andplotted, when the beam stopper and the lateral shield at 90°were used �Fig. 4�b��. In Fig. 5, x-ray measurements per-formed in the XY plane, that is, the plane perpendicular tothe electron beam direction, with and without lateral shield,are reported. The stray radiation equivalent dose values �ex-pressed in units of �Sv /Gy�, measured at 1 m along theelectron beam axis �0°� and orthogonally �90°� for the refer-ence and the clinically most used applicators, are also re-ported in Table III.
Table IV reports values of neutron contamination mea-surements around the 12 MeV LIAC™ for the measurementpoints indicated in Fig. 2. The results are expressed as theneutron dose equivalent per unit of delivered electron dose at
FIG. 5. Stray radiation dose equivalent per unit of electron beam dose de-livered at Rmax ��Sv /Gy�, 12 MeV beam energy, 100 mm diameter appli-cator, SSD 60 cm, 1 m away from electron beam impact on patient in the XYplane with and without the lateral shield.
TABLE III. X ray along e-beam axis 0° and orthogonally �90°� ��Sv /Gy� at1 m of distance.
1001 Soriani et al.: Radiation protection around 12 MeV dedicated IORT accelerator 1001
Rmax �in units of �Sv /Gy�. These measurements, carried outusing the bubble detectors, were performed with an environ-mental temperature ranging from 19 to 25 °C and tempera-ture correction factors ranging from 0.95 to 1.3 were applied.The dose equivalent rate of fast neutrons at the external sur-face of accelerator head ranged from 0.22�0.05 to1.76�0.35 �Sv /Gy. The measured rate varied along thelength of the external surface of the applicator as shown inFig. 2 and Table IV. The highest measured value was foundat the surface of the phantom on the central beam axis �pointE of Fig. 2 and Table IV�, while a lower value was observedon the opposite side under the phantom �point H of Fig. 2and Table IV�. The measurements obtained using the BDTdosimeter are also reported in Table IV. Measurements car-ried out with the mobile beam stopper in the clinical useposition are reported in brackets in the same table. In addi-tion, neutron equivalent dose rate values obtained with theproportional counter are given and agree within experimentalerror with those carried out with the bubble detectors.
IV. DISCUSSION
The typical delivered dose during IORT ranges from 10 to25 Gy. The main purpose of this type of mobile accelerator isto deliver a high dose of radiation immediately after thenumber of clonogenic cells is reduced by surgical removal.The PDD of this machine guarantees an adequate dose de-livery both on the surface and in the depth region whereresidual subclinical disease could still be present. The differ-ence in the PDD curves, with and without the applicator, canbe explained as follows: Electrons scattering off the PMMAapplicator create lower energy electrons, which deposit dosein the first millimeters of the phantom, increasing the surfacedose and moving Rmax toward the surface. Moreover, thePDD curve with the applicator is characterized by a wideplateau, that is, a region in depth with a dose homogeneity
TABLE IV. Neutron dose equivalent per unit of electron dose delivered usingare reported in brackets.
Point of measure�Fig. 2� Description
A Exit window and scattering foil plane outside the coverB Ionization chamber planeC External surface on the applicator middle pointD Phantom surface plane: Outside the applicatorE Phantom surface plane: Inside the applicatorF Phantom surface plane: 15 cm from the applicatorG Phantom surface plane: 15 cm from the applicatorH 15 cm depth �under the phantom�I Floor plane 140 cm from the scattering foilL Phantom surface plane: 100 cm from the applicator
Measurements performedI Floor plane 140 cm from the scatteriL Phantom surface plane: 100 cm from the
within �5%, advantageous from the clinical point of view. A
Medical Physics, Vol. 37, No. 3, March 2010
Monte Carlo simulation has been carried out confirming thisqualitative analysis and these results will be included in afuture paper.
The use of 12 MeV with an 100 mm diameter applicatorwas expected to have higher x-ray contamination values be-cause when electron beam energy and applicator diameterincrease, the bremsstrahlung tail increases, so we used x-rayvalues measured with 100 mm applicator in radioprotectioncalculations to be more conservative. The x-ray contamina-tion of the 12 MeV LIAC™ is lower than the conventionallinac �Fig. 3�. This result has a great importance in theshielding of environmental radiation.
The TVL and HVL of patient scattered radiation, mea-sured at 0° and at 90°, correspond to values generated by aphoton beam with a mean energy of 1.7 and 0.5 MV, respec-tively. On the other hand, the backscatter component mea-sured at 180° is negligible �Table II�. These data are in ac-cordance with Sorcini et al.30 reporting a bremsstrahlungphoton mean energy approximately 1/7 of the incident elec-tron energy.
The x-ray stray radiation values decrease when the anglewith respect to the electron beam axis increases and at 90°there is a tenfold reduction in the exposure �Table III andFig. 4�. Moreover, due to a lower mean energy of the x-raystray radiation a lateral shield of 53 mm of steel and 10 mmof polycarbonate decreases the exposure by a factor of about10. The use of the beam stopper shield decreases the expo-sure along the beam axis by a factor of about 200 and guar-antees the radioprotection of workers in the room below evenwhen we assume a precautionary weekly workload of about250 Gy/week �Table III and Fig. 4�b��.
From a radioprotection point of view, assuming that halfof IORT cases are treated with 12 MeV electrons and theother half with 10 MeV electrons, the calculated equivalentdose at 1 m on the beam axis is 2.44 Sv/y, equal to 0.27 Sv/y
™; values obtained with the mobile beam stopper in its clinical use position
ron equivalent dose per unit of absorbed dose delivered at Rmax ��Sv /Gy�
With the applicator Without the applicatorBDT BD100R BDT BD100R
at 3 m. Considering the attenuation of the beam absorber
1002 Soriani et al.: Radiation protection around 12 MeV dedicated IORT accelerator 1002
described above and that of the floor �assuming standardbuilding materials�, the equivalent dose per year calculatedin the room below becomes 0.86 mSv/y, while �based on the180° value� equivalent dose per year calculated in the roomupper is lower than 0.1 mSv/y.
Regarding the plane perpendicular to electron beam direc-tion, the horizontal plane XY, stray radiation pattern is sym-metric except for the attenuation given by the LIAC™ itself.As long as this attenuation factor is relevant only near theaccelerator, where nobody is allowed to stay during the irra-diation, we have decided not to consider it to be more con-servative. Consequently the stray radiation pattern in thehorizontal plane XY is assumed strictly symmetric in the cal-culations. The equivalent dose at 1 m 90° with respect to thebeam axis is 234 mSv/y, equal to 14.6 mSv/y at 4 m far fromthe LIAC, which is the position of the IORT accelerator con-trol unit in our Institute. Considering the attenuation of thelateral shield and an occupancy factor of 0.5, we have aequivalent dose per year of 0.75 mSv/y. In case of differentlayout and operative conditions, most restrictive workloadvalue needs to be considered.
The presence of the TVL lateral shields generates a conein which the stray radiation is reduced, as shown in Fig. 5.
The maximum value of the dose equivalent rate of fastneutrons at the external surface of the accelerator head wasfound close to the ionization chamber plane �Table IV�. Thehigher measured values �equal to 2.88 and 1.55 �Sv /Gy�were found at the surface of patient simulating phantom onthe central beam axis. It is important to note that at 1 morthogonally and on the floor under the shield, the rate of fastneutrons was negligible.
According to Vanhavere et al.27 the measurements regis-tered by BDT is due to both fast and slow neutrons, and theresponse of BDT to fast neutrons gives a detection rate of10%. Consequently, our results indicate �Table IV� that theslow neutron contamination rate was below the thresholdsensitivity.
Finally, our results are similar to values reported by Loi etal.6 that were carried out around a 12 MeV Mobetron �In-traOp Medical Inc., Santa Clara, CA� and higher than re-ported in our previously published paper3 obtained for a lowenergy model of LIAC™. Our measurements are five timeslower than those reported by Jaradat and Biggs5 for the Me-vatron ME �Siemens, Concord, CA� as expected, due to thelightweight materials used, showing a reduction in x-ray con-tamination tails. Both the LIAC and the Mevatron ME neu-tron contamination rates are lower than the conventional lin-ear accelerators both at 0° and 90°.5
V. CONCLUSIONS
Due to the clinical and therapeutic issues in treating deeptargets and/or overcoming the thickness of biological fluidsduring surgical procedures, the maximum energy of LIAC™has been increased and some critical radioprotection con-cerns need to be carefully addressed. Thus, the x-ray andneutron contamination dose equivalent have been measured
to characterize this new model.
Medical Physics, Vol. 37, No. 3, March 2010
Our data are slightly lower compared to published datafor intraoperative mobile and conventional linacs. Measuredvalues have been used in radioprotection calculations, as-suming a typical workload of 250Gy/week. These resultsconfirm that mobile shielding panels can be used when nec-essary to keep radiation levels around the new LIAC™ be-low the regulatory limits without necessarily providing fixedshielding in the operating room.
ACKNOWLEDGMENTS
The authors wish to acknowledge the kind and meaning-ful contribution from Assunta Petrucci, Ph.D., Medical Phys-ics Department of S. Filippo Neri Hospital in Rome.
1M. Pimpinella et al., “Dosimetric characteristics of electron beams pro-duced by a mobile accelerator for IORT,” Phys. Med. Biol. 52�20�, 6197–6214 �2007�.
2M. D. Mills et al., “Commissioning of a mobile electron accelerator forintraoperative radiotherapy,” J. Appl. Clin. Med. Phys. 2�3�, 121–130�2001�.
3L. Strigari et al., “Radiation exposure of personnel during intraoperativeradiotherapy �IORT�: Radiation protection aspects,” J. Exp. Clin. CancerRes. 23�3�, 489–494 �2004�.
4C. C. Chen et al., “A detailed study on the neutron contamination for a 10MeV medical electron accelerator,” Nucl. Instrum. Methods Phys. Res. A562, 1033–1037 �2006�.
5A. K. Jaradat and P. J. Biggs, “Measurement of the neutron leakage froma dedicated intraoperative radiation therapy electron linear accelerator anda conventional linear accelerator for 9, 12, 15�16�, and 18�20� MeV elec-tron energies,” Med. Phys. 35�5�, 1711–1717 �2008�.
6G. Loi et al., “Neutron production from a mobile linear accelerator oper-ating in electron mode for intraoperative radiation therapy,” Phys. Med.Biol. 51�3�, 695–702 �2006�.
7J. L. Daves and M. D. Mills, “Shielding assessment of a mobile electronaccelerator for intraoperative radiotherapy,” J. Appl. Clin. Med. Phys.2�3�, 165–173 �2001�.
8Council Directive 96/29, EURATOM, “Laying down basic safety stan-dards for the protection of the health of workers and the general publicagainst the dangers arising from ionizing radiation,” 13 May 1996.
9Istituto Superiore di Sanità, “Guidelines for quality assurance in intra-operative radiation therapy,” ISTISAN Report No. 03/1 EN, 2003.
10A. Soriani et al., “Setup verification and in vivo dosimetry during intra-operative radiation therapy �IORT� for prostate cancer,” Med. Phys.34�8�, 3205–3210 �2007�.
11Comitato Elettrotecnico Italiano �CEI�, “Norme particolari di sicurezzaper gli acceleratori medicali di elettroni nella gamma da 1 MeV a 50MeV,” CEI EN 60601–2-1, Class CEI 62–35, CT 62, 2000.
12T. P. Wangler, “Introduction to linear accelerator,” Los Alamos NationalLaboratory Report No. LA-UR-93–805, 1993.
13http://iba-dosimetry.com/Detectors.704.0.html14http://www.ptw.de/roos_electron_chamber0.html15International Atomic Energy Agency, “Absorbed dose in external beam
radiotherapy: An international code of practice for dosimetry based onstandards of absorbed dose to water,” IAEA Technical Report Series No.398, 2001.
16F. Di Martino, M. Giannelli, A. C. Traino, and M. Lazzeri, “Ion recom-bination correction for very high dose-per-pulse high energy electronsbeams,” Med. Phys. 32, 2204–2210 �2005�.
17R. F. Laitano et al., “Charge collection efficiency in ionization chambersexposed to electron beams with high dose per pulse,” Phys. Med. Biol.51�24�, 6419–6436 �2006�.
18http://www.elimpex.com/products/survey-meter/victoreen451b.pdf19NCRP, “Radiation protection design guidelines for 0.1–100MeV particle
accelerator facilities,” NCRP Report No. 51, 1977.20NCRP, “Neutron contamination from medical electron accelerators,”
NCRP Report No. 79, 1984.21
NCRP, “Structural shielding design and evaluation for megavoltage x-
28ICRP, Recommendations of the international commission on radiologicalprotection, ICRP publication 60 �1990�.
29ICRP, Conversion coefficients for use in radiological protection againstexternal radiation, ICRP publication 74 �1996�.
30B. Sorcini, S. Hyodynmaa, and A. Brahme, “Quantification of mean en-ergy and photon contamination for accurate dosimetry of high-energyelectron beams,” Phys. Med. Biol. 42, 1849–1873 �1997�.
